As people wait to take the Binax rapid test at 24th St. Mission BART Station Plaza, they fill out online surveys that question travel and behavioral habits on their phones. Photo taken by Annika Hom on Jan. 10, 2021.

The most powerful new weapon in the city’s war on Covid-19 is a mid-sized daily testing site at the 24th Street BART Plaza. 

For 15 days in January, UCSF researchers and the Latino Task Force tested some 600 people a day. Each day, the site found anywhere from 16 to 46 percent of all positive cases in the city, according to data from the Department of Public Health. That efficiency would be akin to a small crew of miners competing against an army of pitmen, and day after day extracting 16 to 46 percent of the gold. 

Moreover, because the 24th Street site uses rapid tests — testing that catches virus carriers with high viral loads within 15 minutes of being tested — it was able to get those who tested positive into quarantine on the same day. 

“The 24th Street testing is actually getting infected people off the street,” said Peter Khoury, a data scientist who has followed the Mission testing campaigns. “Not only did it find them at a higher rate, but it found them sooner” than conventional testing.

Khoury calculates that the 24th Street BART Plaza, which continues its testing this month, is 2.3 times more effective than the Embarcadero site, and 1.6 times more effective than all SFDPH sites.

Take Sunday, Jan. 17, when some 3,450 people were tested across the city. The BART site’s 713 tests represented 20 percent of the citywide total, but its 66 positive cases accounted for 45 percent of the city’s 145 positives.

On Tuesday, Jan. 19, when the city conducted a total of 8,515 tests, the BART site tested only 567 people, less than 7 percent of all tests. Nevertheless, it managed to find 70 cases, or 26 percent of the city’s 266 total positive cases. 

To Dr. Diane Havlir, the lead researcher on the project and chief of UCSF’s Division of HIV/AIDS, Infectious Disease, and Global Medicine, the 24th BART Plaza’s success speaks to two factors: convenience and location.  

In its favor, the site is open four days a week (now Sunday through Wednesday) and the hours are regular (9 a.m. to 2 p.m.). It is also readily accessible to residents of neighboring communities where, like the Mission, the prevalence of the virus is high.  

Or, as Havlir’s aide de camp, Diane Jones, a site leader, put it, “This is where the virus is living.” 

Moreover, thanks to rapid testing, the site has acquired a reputation for responding to a positive result with same-day assistance, and follow-up from the Latino Task Force’s wellness teams. And, if the positive person lives with three or more people, UCSF Researcher Carina Marquez will send a team to test everyone in the household. 

That approach is critical to the city’s strategy in controlling the virus. While the excitement is around vaccines — the UCSF/Latino Task Force collaborative has its vaccine site on 24th and Capp streets — researchers warn that it will take months before vaccines become ubiquitous enough to create immunity in the community. 

The announcement on Sunday that the city would pause vaccinations at Moscone Center, and wait until Friday to re-open the City College site, was another reminder of testing’s importance in keeping the virus under control.

“In the meantime, we need to minimize the harm of this virus with a public health response of ‘test and respond’ approach to interrupt transmission and support those who are infected,” wrote Havlir in an email.

While no test is perfect, the ability to detect the virus quickly and get someone quarantined the same day is “a powerful tool,” said Havlir, that “finds those with the highest viral levels, [and] allows immediate action to interrupt the chain of transmission.” 

Havlir wrote that she could see more use of rapid testing in the community as well as in densely populated settings such as shelters and schools.

Indeed, the city has been watching closely. A spokesperson for the health department wrote in an email that “We are working on rolling out Binax (the rapid tests) at our community clinics.”

And, the city’s own testing strategy has improved considerably since the fall, when it provided most of its testing resources in the least impacted neighborhoods. 

Testing during one week in January

To understand the health department’s evolving testing strategy, Mission Local looked at testing during the week of Jan. 11 to 17. We started with that week because we had data from the Department of Public Health popup sites for those dates, and the health department  filled out the picture by giving us aggregated data from its other sites: the Embarcadero, the Alemany Farmers market, and the so-called alternative sites, which are seven neighborhood clinics in the southeast that performed a total of 1,383 tests. (The health department did not break out how many tests were done at each of the clinics.)

During this period, the 24th Street BART site, which administered 16 percent of all the tests except those by private insurers, found 23 percent of the 1,344 positives found by DPH and UCSF/LTF. 

The stars of the city-run sites were clearly the so-called alternate sites or neighborhood clinics. The seven sites, all in impacted neighborhoods, represented only 7 percent of the DPH and UCSF/LTF tests, but found 16  percent of the positive cases. 

Clearly, the city’s clinics and the 24th Street BART station delivered more bang for the buck. 

But the city’s decision in mid-November to close SoMa after months of low positivity and move those tests to the Alemany Market has also meant the city is identifying more covid-positive residents. During the week of Jan. 11, Alemany accounted for 21 percent of all DPH tests, compared to 56 percent at Embarcadero, but it caught 27 percent of all DPH positives compared to 37 percent at Embarcadero, making it twice as effective. 

The tests done by UCSF/LTF at the BART Plaza and Alemany accounted for 17 percent of the tests, and 21 percent of the positives. The BART Plaza’s share was 16 percent of the tests, and 23 percent of the positives.

Moreover, like the BART Plaza, Alemany proved such a popular drop-in site that the city no longer takes online appointments. That success underscores what the UCSF/LTF site has long demonstrated: make testing low-barrier, regular, and close to impacted residents, and those populations will use it.

It’s unclear why the popups have consistently found fewer positives, even though they are located in impacted communities. Researchers said the results could simply reflect the sites attracting fewer vulnerable residents. And that, in turn, is a problem of outreach, and perhaps the infrequent nature of the testing operations. 

For the week that we analyzed data, the health department concentrated more than half its tests — 55 percent — at the Embarcadero, where the positivity rate was lowest. But when Mission Local began writing about testing disparities in September, the health department  had 70 percent of its testing resources at the Embarcadero.

A spokesperson for the city said that 700 of its 1,700 tests have been reallocated to the impacted communities, but other than Alemany it is unclear where those tests are going. 

There’s no doubt that the city has increased testing at its popups since it started this summer. But in the last several months, the tests have held steady. For example, the Bayview’s three popup sites, from Nov. 30 to Dec. 4, per health department slides, reported 480 tests; that number jumped to 575 the week of Jan. 11, and then fell back down to 474 the week of Jan. 17.

However, Alemany and the pop-ups have increased the testing rates in the southeastern neighborhoods.  Bayview Hunter’s Point, for example, has gone from testing 124 residents per 1,000 in September to 285 residents per 1,000 residents this month. Visitacion Valley has gone from testing 89 residents per 1,000 in September  to 255. 

But even at these rates, the neighborhoods remain far below the 700-plus per 1,000 in places like Mission Bay and Japantown. Most importantly, the positivity rates in Bayview and Visitacion Valley remain the highest in the city. The 16 percent positivity rate at the Department of Public Health alternate clinics reflect neighborhoods in crisis. 

When asked about the health department’s strategy, a spokesperson responded, “We continue to reallocate as needed. The City’s flexible and strategic testing strategy allows us to deploy resources to hard-hit communities to ensure equitable and effective testing that helps us slow the spread and save lives.”

UCSF researchers say it’s all a matter of acting on the data and putting resources where they are needed. Already, the Alemany site is paying off. More tests and rapid tests at the alternate sites could also keep more of those with the virus isolated.

The ultimate goal, as Havlir’s team has shown at BART, is to get virus carriers quarantined on the same day they receive test results. 

Dr. Kirstin Bibbins-Domingo, chair of epidemiology at UCSF and Vice Dean for Population Health and Health Equity, said  “This all comes down to resources and a willingness to accept the fact that you have to put resources where there is need.” 

It is not just that  “these communities are suffering disproportionately,” she added. “Any strategy to reopen the city depends on bringing the transmission down in the communities where infections are the highest, because we are all interrelated — all parts of town.”

Annika Hom contributed to this reporting.

This article was produced as a project for the USC Annenberg Center for Health Journalism’s 2020 Data Fellowship.

If you have a Covid-related story you would like to share, please be in touch

Follow Us

Founder/Executive Editor. I’ve been a Mission resident since 1998 and a professor emeritus at Berkeley’s J-school since 2019 when I retired. I got my start in newspapers at the Albuquerque Tribune in the city where I was born and raised. Like many local news outlets, The Tribune no longer exists. I left daily newspapers after working at The New York Times for the business, foreign and city desks. Lucky for all of us, it is still there.

As an old friend once pointed out, local has long been in my bones. My Master’s Project at Columbia, later published in New York Magazine, was on New York City’s experiment in community boards.

Right now I'm trying to figure out how you make that long-held interest in local news sustainable. The answer continues to elude me.

Join the Conversation


  1. Interesting that even with all that data, quarantining and outreach, the Latinx positivity rate is still so high, nearly as high as it was 6 months ago. Makes me wonder what we more needs to be done in terms of tracing, quarantining, financial support and community messaging to significantly drive down infection rates in the community. While many of the issues are structural (essential workers, can’t afford to miss work, crowded housing, etc.) , behavioral issues seem to contribute too. I live in the very diverse Outer Mission, and Latinx (mostly men) are by far the largest group I see gathering without masks, playing soccer and baseball and hanging out with friends and family.

    votes. Sign in to vote
  2. Hi,
    What about antibodies testing for COVID-19, is one of the key factors contact tracing for covid 19. I know that health provider like Kaiser Permanent isn’t doing the antibodies testing because they feel there isn’t enough research, which say they need to implement as part at their testing facilities.
    Be Safe,

    votes. Sign in to vote
  3. Why are they still doing so many tests at the Embarcadero site? It’s a total waste of resources. They have consistently tiny numbers of positive tests and yet that’s where 55% of the testing is still being done. In other cities testing is concentrated where the cases are highest, in crowded, poorer neighborhoods, with more front-line essential workers. I contend that the powers that be do not want to know the true prevalence of covid in SF because it would impact their reopening plans and they also don’t want to commit resources to more contact tracing and quarantining.
    Look how long it took them to start testing in the Mission–more than 7 months into the pandemic– and only after lots of community pressure and outrage. And that was once a week at the Alemany market until December or so. It only stepped up when there was a partnership with UCSF and the Latino Task Force.
    It’s beyond enraging.

    votes. Sign in to vote
  4. Do you have any information on the rate of false positives in the rapid test result data? A high false positivity rate would skew the data as well as needlessly force many people into quarantine creating financial and emotional problems. I have heard from a nurse at SF General that ~6/10 of patients who receive a rapid test positive result during admission in the emergency department end up testing negative with the more accurate follow-up PCR tests? Are there any official statistics available?

    votes. Sign in to vote
    1. Good questions Mark. The researchers have been careful in testing this first back in September at 16th Street and then again during the holiday testing in which they used both the rapid test and the PCR. Here is a story from the Thanksgiving testing
      As you can see PCR is most sensitive but the researchers feel that the Binax gets positive results when the person is infected has a higher viral load. They are very intentional in getting people into quarantine as quickly as possible and the rapid test has allowed that. Here is another story in which researchers talk about the reasoning.

      votes. Sign in to vote
    2. Mark the protocol put in place in January was only after a careful study back in October and in Thanksgiving in which Diane Havlir and Unidos en Salud ran both rapid and PCR tests together. They confirmed in both studies that the rapid tests did not produce false positives. The rapid tests caught a few less people than the super sensitive PCR tests but they concluded that those people probably weren’t contagious and the PCR tests were just catching inactive virus.

      votes. Sign in to vote
  5. The key here is not just the (relatively) higher test positivity at the BART station; finding positives doesn’t mean anything if you’re not stopping transmission from those positives. It’s that they’re able to quarantine people same day. There was a study done on contact tracing in the city through July and at each step, you have losses; maybe you can contact 80% of people who test positive. Maybe 80% of those name a contact. Maybe you reach 75% of those, and of those 50% agree to take a test. And then about 11% test positive — some of whom may have been infected in the 2-3 days for that process to finish up. So all that contact tracing is shaving off maybe 2% or 3% AT MOST of the transmission occurring in the city. Wherever the sites are located, they need to roll out dual-tests (rapid+PCR) at EVERY site, and have on-site contact tracers. That cuts the losses in the first 2 steps, and reduces how many people in that last step test positive. You can potentially triple the impact of contact tracing and actually make a big difference in transmission.

    votes. Sign in to vote
    1. Yes, it is so true about contact tracing. What’s been remarkable about the 24th Street site is how quickly they do get the person and a household the information needed to quarantine. The city takes over on the contact tracing beyond the household, but now that cases have come down, it’s possible the health department is getting to more people. That is something we should check on. Best, Lydia

      votes. Sign in to vote
Leave a comment
Please keep your comments short and civil. Do not leave multiple comments under multiple names on one article. We will zap comments that fail to adhere to these short and very easy-to-follow rules.

Your email address will not be published. Required fields are marked *